Ahsdown MAG 600H Main fuses blows

[Pizzaiolo]

Hi everyone! My first post on this forum while I repair a bass amp for a friend.

Maybe a little introduction to the story could help to understand the blowing fuses.

Thoses amps seems blowing because bad thermal junctions causing overheating. In fact the thermal paste was complete dry on the NTC and no longer in contact with the heatsink. In addition to that all transistors were shorted causing main fuse to melt.
I also thought 330mOhm resistors were shorted but I realize that my tools are not enough precise to make proof of that.

I did replace all the transistors and resistors too. I did also put a clip to hold the NTC on the heatsink and put new thermal paste everywhere.

On the first try it did almost work. No blowing fuse but very pronounced noisy/distorted sound. My first suspect was bad resistor type because I took carbon ones. So replaced them with metal-film ones. No sound. After a bit of head-scratching I realized that I omitted the "0" before the "R" on the schematic meaning that the resistors were 10x grater than normal. OK! new ones. Nice Clean Big Fat Sound!!! Oh yeah baby!!! Jammin jammin jammin FUSE BLOWN; tears... I took it back home and after blowing some more fuses I fried the transistors again.
First there were only one pair of them shorted out on every pin and after removing them the sound was back. Sadly that didn't last very long when I turned the volume up the fuses blow again and the remaining transistors died shorted on all pins.

No I am at the exact same point here I started making me thinking of the process of resolving this issue. I could just have made it worse and I am at the limit of my understanding of the circuitry. I barely know the basic principle of this type of amp. What I see as an H-bridge class D amp. But I am not sure. If so, there is a control circuitry that may have been affected by those over-currents.

I am not sure very sure what to do next and I hope someone could suggest some components to test or an other methodology to apply. I read about light-bulb limiter. I don't have the right components at this time.

What I will do next unless someone tell me not to is to test power supply on the remaining components. But I am not sure what to expect. And what risk for other components if I turn it on without the transistors. I doubt there is any but at this point i prefer keep thinking about it and look for help. I could also track the signal path by an oscilloscope if that could also be a good solution.

I am also aware of detailed functional documentation about this type of amp. What I have found was not very revealing to me.

Thank you for reading hope it was not too much of a pain!

I replaced R2, R3, R6, R20, R33, R34 and TR1, TR4,TR6, TR7, TR13, TR19, TR21, TR22
Schematic:
https://music-electronics-forum.com/...0&d=1394892610

[Enzo]

Hi

First DISCONNECT the speaker. DO not connect a speaker or load until the amp is stable and not making DC. ONLY THEN connect a speaker.

Check the output for DC voltage, there should be none. A few millivolts is OK.

Look up "light bulb limiter", then make one and use it. Every time you blow another fuse it stresses the amplifier more.

This is a basic conventional amp circuit, not class H or D. It has just the two main power rails.

Do not worry about the value o the 0.33 ohm resistors by the output transistors. They will either be open or they will be OK. They won't be wrong value.

You need to monitor current. Either the mains draw, or even just by the voltage drop across any of those 0.33 ohm resistors.

[Pizzaiolo]

Hi Enzo,

I will build that light bulb limiter when I have found one.

So if I understand you right I should replace all the faulty transistors and measure voltage drop accross 0.33ohm resistor. Will I do that with a signal at the input or without? And from there faulty control circtuitry could be tested?

Best,
David

[Enzo]

No signal until we know the amp is stable and not producing DC, and that includes idle current. If the amp cannot even sit there at idle, is surely will not reliably amplify anything.

[Mick Bailey]

Dave Green at Ashdown recommends replacing the 2SA1695 with 2SA1294, and the 2SC4468 with 2SC3263 in these amps whenever the originals short. While the amp is without its output transistors it's also convenient to power it up off the limiter and see if there's any unusual voltage on the base connections.

Also, I've had transistors blow straight away with no other fault and you have to be sure they're not a repackaged or fake lower spec device.

[Pizzaiolo]

Hi Mick,

Thank you I will follow this advice as I have not yet ordered new ones. I've got mine from Reichelt.com what seems to be a serious distributor of electronics to mee. They are from ISCSEMI. Should be ok to me.

I have found a lightbulb but not yet build the limiter because there is actually no more short and before to order new components I wanted to check some other transistor. I have desoldered and measured TR10 and compared the values measured with a fresh one I have in stock. I didn't measured open circuit on any pin so replaced it straight. The same for TR2. Then I went to TR3 but I have nothing to compare with so I'm asking you guys what you think.

When I put (+)probe one the base and (-)probe on collector I am measuring about 18.3MOhms and 19.4MOhms on emitter. When inverting probe (-)probe on base, I measure open circuit between base-collector and base-emitter. Collector and emitter are open both direction. Is this meaning something to you?

EDIT: I have not found 2SC2363 but only 2SC3263 maybe you have inverted the 2 and 3? https://www.reichelt.com/ch/en/trans...88.html?&nbc=1
This is the only pair matching 2SA1294.

[Mick Bailey]

Sorry - I transposed the digits - I;ve edited my post and the part you've quoted is correct.

Set your DMM to diode mode as a more reliable way to test the transistor;

https://engineeringtutorial.com/test...al-multimeter/

[Pizzaiolo]

Yep amazing Mick thanks that's the kind of information I needed! Very useful link!

So in fact some of them either half-open or shorted through. I have ordered enough to replace them all and changed the jamicon filtering caps to elna ones. Would be adding an extra 4700uF pair capacitor be worth the 10$?

And next will be to test the thing with a limiter wired in. And if there's no short I should take some measurement. First on the speaker output for any DC voltage offset, right? If there is some that could mean the bias is not right, yes? The polarity of the measurement should also indicate which side? Is it that simple?
I'm not really Sure. The polarization doesn't look very simple nor obvious. Black magic at this point. But I think I must focus on the end result and isolate that part. Which means that the next first thing would be the voltage between collector and emitter resistance. I try but I cannot figure out what I should expect on PIN2 of TR13 and PIN3 of TR5 at idle state. For example. Don't even know if it is a good one. Anyway... Thank's!

EDIT: Power supply looks good so far.

[Mick Bailey]

My thought with repairing an amp is to repair it, not modify it along the way. Say you add capacitance and the rectifier fails - then you don't know if the added capacitance stressed the rectifier, or there is another fault. So then you remove the additional capacitance, and replace the rectifier. Back to where you started. A modify-as-you-fix approach gets you into trouble. Even seasoned pros don't redesign an amp as they're fixing it. Replacing all the transistors is something I've never needed to do. These amps are pretty robust and it's unusual for output shorts to take out the drivers. When you find a faulty part, mark on the schematic in red where it is and determine the failure path. Very often you can see if something's failed and how the particular failure mode affects other connected parts. Replacing a good part introduces the possibility of the replacement part being faulty. Always be sure that a part has failed or highly suspect before replacing it.

Take your time and reason things out. There's more satisfaction in looking at a board with hundreds of components and identifying the single part that's causing it to fail, than replacing dozens of parts and it still not work.

Excessive DC offset on the speaker output is a bad sign. Biasing does not affect this to a significant degree because the bias affect the current draw equally on both sides of the push-pull. What you get with excessive bias is more current draw at idle than you want, resulting in the amp getting warm without any signal. DC offset means that one side of the P-P is pulling more towards one power rail. That means one side is being turned on more than the other. Transistors are controlled by their bases, so look for unwanted or high base voltage. If you replace the outputs and see a DC offset but everything else checks out OK, remove the drives which should take away the base voltage. If it doesn't you've simplified the problem and eliminated everything else other than the outputs and power supply.

[Pizzaiolo]

Take your time and reason things out. There's more satisfaction in looking at a board with hundreds of components and identifying the single part that's causing it to fail, than replacing dozens of parts and it still not work.

I must agree with this and above!

But for this particular unit I must say that I had to improve some things. Because of the bad thermal junctions with the overheating protection circuitry. I added a clip to firmly hold the NTC in place and in contact with the heat-sink as an other user on this forum did. Thus avoiding the same problem happen again (Hope so!). Now the new transistors with a higher rated current is also good. There is maybe also better thermal junctions and dissipation capabilities so this goes also in the same direction. Modifying - Improving - an existing design. So there are always exceptions!

By the way I dono how more filtering capacity could stress a rectifier... maybe you can tell me? When I see a little 2 amp PSU with the same amount of capacity I am a bit surprised. But maybe is ripple also of little consideration in this type of amp? I want to learn amore but that's maybe not the subject! (:

Anyway, thank you guys for advice and help! Hope the new parts will arrive tomorrow and get some time to replace them! (:

[Pizzaiolo]

Hi!

Problem solved! A few mV DC at speakers. Everything working fine!

Thank's a lot for the good advices.

[Mick Bailey]

By the way I dono how more filtering capacity could stress a rectifier... maybe you can tell me? When I see a little 2 amp PSU with the same amount of capacity I am a bit surprised.

Does the 2amp PSU have the same rail voltage as the Ashdown, is it a split-rail supply, are the rectifiers the same? Inrush current stresses a rectifier, and the larger the capacitance the greater the chance of the PSU charging asymmetrically, causing loud thumps when turning on or off. When you have a fault condition the greater energy stored in larger caps can do more damage.

...so this goes also in the same direction. Modifying - Improving - an existing design. So there are always exceptions!.

Dave Green is one of the amp designers at Ashdown, and highly regarded within the industry. When the manufacturer recommends a change, that's not the same as someone repairing the amp making a design change.

Jack Darr has a good handle on repair work as opposed to design;

"We technicians have another kind of problem. Our work is not design in any sense of the word. We are maintenance men; we work with units that are already built and at one time worked! We have to use an entirely different kind of thinking. Notice the key word: thinking! When we start a job we have the benefit of this one priceless bit of data: "This thing did work at one time or it couldn't have been sold!" So our job is not to redesign it, but simply to make it work again as well as it did when it was new. You'd be surprised how many technicians forget this. They attack a problem as if it were a design problem. They forget that what they're looking for is a bad part, a "bug" in a circuit that definitely did work at one time. If we remember this, it will always give us a kind of handle for the problem"